Download Temperature control of electromigration to form gold nanogap

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Islanding wikipedia , lookup

Electrical ballast wikipedia , lookup

Three-phase electric power wikipedia , lookup

Rectifier wikipedia , lookup

Electrical substation wikipedia , lookup

Power engineering wikipedia , lookup

Negative feedback wikipedia , lookup

History of electric power transmission wikipedia , lookup

Voltage regulator wikipedia , lookup

Current source wikipedia , lookup

Power electronics wikipedia , lookup

Opto-isolator wikipedia , lookup

Buck converter wikipedia , lookup

Nanowire wikipedia , lookup

Control system wikipedia , lookup

Stray voltage wikipedia , lookup

Rectiverter wikipedia , lookup

P–n diode wikipedia , lookup

Surge protector wikipedia , lookup

Metadyne wikipedia , lookup

Switched-mode power supply wikipedia , lookup

Lumped element model wikipedia , lookup

Ohm's law wikipedia , lookup

Voltage optimisation wikipedia , lookup

Resistive opto-isolator wikipedia , lookup

Mains electricity wikipedia , lookup

Alternating current wikipedia , lookup

Thermal runaway wikipedia , lookup

Transcript
APPLIED PHYSICS LETTERS 87, 263101 共2005兲
Temperature control of electromigration to form gold nanogap junctions
G. Esen and M. S. Fuhrera兲
Department of Physics and Center for Superconductivity Research, University of Maryland,
College Park, Maryland 20742-4111
共Received 3 August 2005; accepted 25 October 2005; published online 19 December 2005兲
Controlled electromigration of gold nanowires of different cross-sectional areas to form nanogap
junctions is studied using a feedback method. A linear correlation between the cross-sectional area
of the gold nanowires and the power dissipated in the junction during electromigration is observed,
indicating that the feedback mechanism operates primarily by controlling the temperature of the
junction during electromigration. We also show that the role of the external feedback circuit is to
prevent thermal runaway; minimization of series resistance allows controlled electromigration to a
significant range of junction resistances with a simple voltage ramp. © 2005 American Institute of
Physics. 关DOI: 10.1063/1.2149174兴
Electromigration 共EM兲, the electrical current-induced
diffusion of atoms in a thin metal film, is important as a
failure mode in integrated circuit interconnects.1 Failure of a
narrow metal wire due to EM has also recently been utilized
extensively to prepare stable electrical contact pairs with nanometer separation for single molecule electrical
experiments.2,3 It has been suggested2–4 that the dominant
failure mechanism in such electrically stressed gold nanowires is thermally assisted EM. Recently, feedback schemes
for controlling the rate of EM were shown to allow control
over the final junction resistance 共and presumably gap size兲
in such nanogap junctions.5,6
The empirical formula for the median time to failure
MTTF= Aj−neEa/kT, known as the Black equation7 has been
employed to determine interconnect reliability, where A is a
sample dependent constant, j is the current density, n is an
exponent empirically found between 1 and 7,8 Ea is an activation energy for atomic motion, k is the Boltzmann constant, and T is the temperature. However, in the formation of
nanogap junctions by the EM of a short nanowire, the current
and temperature 共due to Joule heating兲 are both changing
rapidly. Understanding the role of current and temperature is
then critical to the design of circuits to produce nanogap
junctions in a controlled manner.
In recent works,5,6 it was proposed that controllable EM
occurs at constant applied power. Furthermore, it was stated5
that electromigration is triggered at a constant temperature.
In this letter, by experimenting on different nanowire geometries, we confirm that EM occurs at constant temperature
independent of geometry, and show that the feedback mechanism works primarily through controlling the temperature of
the electromigrating junction. Furthermore, we estimate the
temperature of the junction during EM to be only a few
hundred Kelvins, low enough to allow the study of many
molecular adsorbates without desorption or dissociation of
the molecules. We also show that the role of the external
feedback circuit is to prevent thermal runaway in the junction, allowing control in the region in which the temperature
increases with increasing junction resistance. This region can
be minimized by minimizing series resistance in the circuit,
a兲
Author to whom correspondence should be addressed; electronic mail:
[email protected]
allowing significant control of the EM with a simple voltage
ramp.
Our devices are fabricated in two steps using conventional electron-beam lithography 共EBL兲 and lift off on SiO2
共500 nm兲 / Si substrates. We first fabricate thin gold lines
共with no adhesion layer兲 of width 40 nm to 100 nm and
thickness 15 to 30 nm. In the second EBL step, we deposited
5 nm Cr and 70 nm Au to form contacts and bonding pads. A
scanning electron microscope 共SEM兲 micrograph of such a
device is shown in the inset of Fig. 1.
To control the EM process, we used a computercontrolled feedback scheme similar to Ref. 5, consisting of
the following steps: We first measure a reference conductance value at a voltage of 100 mV. We then increase the
voltage until the conductance drops by a set fraction 共typically 2–5%兲 of the reference conductance value. At this
point, the voltage bias is ramped down by 50 to 100 mV 共at
FIG. 1. 共Color兲 Current vs bias voltage during the feedback-controlled electromigration of an Au wire at T = 1.3 K. Part A is a smooth curve indicating
than the EM has not begun, whereas in Part B the resistance of the line
increases irreversibly due to EM. Both Parts A and B are recorded in a single
voltage biasing process, producing a final resistance of ⬃120 ⍀. At this
point, the voltage was reduced to zero for some time. When the bias process
was restarted in C, the wire resistance is the same, demonstrating that the
EM process may be frozen by turning off the voltage. The inset shows the
SEM micrograph of one of our devices. The scale bar in the inset is 2
micrometers long. Arrows indicate the progression of the curve.
0003-6951/2005/87共26兲/263101/3/$22.50
87, 263101-1
© 2005 American Institute of Physics
Downloaded 21 Mar 2010 to 132.66.144.233. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
263101-2
G. Esen and M. S. Fuhrer
Appl. Phys. Lett. 87, 263101 共2005兲
FIG. 3. Power dissipated in the junction during electromigration vs nanowire cross-sectional area. The power is the average power in the region of
near-constant power seen in Fig. 2.
FIG. 2. Power dissipated in the junction PJ vs the voltage drop at the
junction VJ. The irreversible change in resistance due to EM starts at the
point labeled 共a兲. Inset shows the corresponding current vs bias voltage data.
The starting nanowire has dimensions 830 nm long⫻ 60 nm wide
⫻ 25 nm thick; the length and width of the nanowire is determined using
SEM, and thickness by quartz crystal monitor during gold film deposition.
a rate of 50 mV/ s兲 and a new reference conductance value is
measured. We repeat this process until the desired conductance is reached. We performed our experiments in a gas
flow 4He cryostat at substrate temperatures from 1.3 K to
room temperature.
A representative current versus voltage 共I-Vbias兲 curve
taken through a feedback-controlled EM process for one of
the devices is shown in Fig. 1. The data labeled A show a
smooth I versus Vbias curve indicating that EM has not begun
in the gold wire. Although there is a resistance increase with
increasing bias in A, we found that if we stop the voltage
bias in Region A, this resistance increase is reversible. Such
a reversible resistance increase shows that the gold wire
heats up before EM begins. The data labeled B show that
after this initial heating, the gold wire begins to change resistance irreversibly due to EM. The data labeled C show
that one can stop and restart the voltage bias before the gold
wire totally fails. I-Vbias curves of two bias processes perfectly match each other indicating that in the second biasing
process the gold wire first heats up to the temperature where
significant EM takes place, and then EM restarts.
We now discuss the temperature of the wire during EM.
We assume the total resistance of the circuit R is the sum of
two resistors, RL, the lead resistance 共equal to the total measured resistance at low bias兲 and RJ, the resistance of the
“junction,” the weak spot formed in the wire by EM; i.e.,
RJ = 0 initially. The power dissipated in the junction is then
PJ = I2RJ. Note that RJ includes a contribution from the resistance change of the leads upon heating; and heating due to
the resistance of the nanowire itself is ignored. Hence, PJ is
a rough estimate of the power that is heating the junction, but
should be valid when RJ is significantly non-zero.
Figure 2 shows PJ versus VJ = IRJ, where the inset graph
is the corresponding I-Vbias curve of the data. After the junction begins to increase resistance due to EM 关Point 共a兲兴, the
power dissipated in the junction reaches a relatively constant
value. Constant power dissipation in the junction is observed
over an order of magnitude of junction voltage.
Figure 3 shows that the power dissipated in the junction
during EM is proportional to the nanowire area. Since the
thermal conductance of the wire is proportional to the crosssectional area, the power required to maintain a given wire
temperature should also be proportional to area. Thus, we
conclude from Fig. 3, and the relatively constant power observed in Fig. 2, that the EM rate is dominated by temperature, and the feedback scheme operates to control the wire
temperature. This conclusion is reasonable from the Black
equation, given that temperature enters exponentially, and
voltage 共through current density兲 only algebraically.
We now estimate the temperature of the junction TJ during EM. Here, we neglect the heat conduction to SiO2 substrate by considering the relative magnitude of the thermal
conductivity of gold and SiO2 共Ref. 9兲 and consider the contacts as infinite heat sinks at T = 1.5 K. The temperature at the
midpoint of a wire with uniform power generation over its
volume is TJ = PL2 / 8V␬ where P, V, L, and ␬ are, respectively, the total power generated in the nanowire, the volume
and length of nanowire, and the thermal conductivity of
gold.10 For the wire in Fig. 2, the maximum power is estimated as ⬃0.67 mW 共including PJ at point 关Fig. 2共a兲兴 and
additional power generated due to the estimated resistance
11 ⍀ of the nanowire at T = 1.5 K兲; using the thermal conductivity of gold as 艌320W / mK 共the room-temperature
value for bulk gold兲 we estimate TJ 艋 145 K. If we consider
instead that all of the power is being generated at the center
of the nanowire and carried out to the leads by the nanowire,
then TJ = PL2 / 4V␬ 艋 290 K; which is still low enough to allow the study of many molecular adsorbates without desorption or dissociation of the molecules.
We now discuss why the feedback process is feasible.
The thermal time constant11 ␶th = L2␳C p / ␲2␬, where C p is the
specific heat, and ␳ is the density, is less than 1 ns in our
wires; much faster than our external feedback circuit. This
suggests that the EM process itself must occur very slowly. If
this is the case, is feedback needed at all? i.e., could Vbias
simply be turned to zero at the desired R? The answer is
no—at constant junction temperature 共i.e., following the
I-Vbias curve in, e.g., Fig. 1兲 I, and hence R, is multiple valued at a given Vbias. Stated another way, the change in temperature with junction resistance dTJ / dRJ during EM must
be negative to prevent thermal runaway once EM begins.
Assuming that TJ is proportional to I2RJ, for our simple se-
Downloaded 21 Mar 2010 to 132.66.144.233. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp
263101-3
Appl. Phys. Lett. 87, 263101 共2005兲
G. Esen and M. S. Fuhrer
FIG. 4. 共Color兲 Current vs bias voltage during the EM of three similar gold
wires 共600– 700 nm long⫻ 40 nm wide⫻ 15 nm thick兲 at T = 1.3 K. For the
blue and red curves, the external feedback is turned off at the points marked
by the blue and red arrows respectively. For the black curve, no external
feedback was used.
ries circuit model dTJ / dRJ ⬍ 0 implies RJ ⬎ RL, corresponding to the stable branch of the I-Vbias curve, where
dVbias / dR ⬎ 0 during EM 共the voltage increases as EM
progresses兲.
Figure 4 illustrates this instability. We turned off the
feedback at various points during the EM process, solely
ramping the voltage upwards at a fixed rate. The red curve
shows the feedback turned off while on the stable I-Vbias
branch 共positive dVbias / dR兲; the current decreases smoothly
with increasing voltage from this point. However, when the
feedback is turned off on the unstable I-Vbias branch 共negative dVbias / dR; blue and black curves兲, the current drops rapidly to the stable branch at the same Vbias. Thus, the feedback
scheme is only necessary to produce final resistances RJ
⬍ RL; with suitable circuit design 共minimization of RL; i.e.,
short nanowires with highly conducting leads兲 small final
RJ’s may be produced using a simple voltage ramp. Note that
in some circuits 共e.g., Fig. 1兲 the stable I-Vbias branch does
not extend beyond the unstable branch; in such cases, a
simple voltage ramp causes abrupt failure of the wire by
melting 共as observed via post-mortem SEM兲, resulting in
large 共⬎10 nm兲 gaps.
In conclusion, we performed controllable EM on nanowires with different cross sections. We found that the average
power dissipated in the junction during EM increases linearly with the area of the junction indicating temperature
control of the process and confirming that the mechanism is
thermally assisted EM. Using the maximum power dissipated in a typical device, we estimate the junction temperature during EM performed at T = 1.5 K to be only a few
hundred Kelvins. We also note that the role of the feedback
process in controlling EM is to prevent thermal runaway in
the region of positive dTJ / dRJ. This region can be reduced
by reducing the series resistance in the circuit, allowing controlled EM with a simple voltage ramp.
This work has been supported by the U.S. Department of
Energy under Grant No. DE-FG02-01ER45939 and by the
NSF-Materials Research Science and Engineering Center under Grant No. DMR-00-80008.
J. R. Lloyd, Semicond. Sci. Technol. 12, 1177 共1997兲, and the references
therein.
2
H. Park, A. K. L. Kim, J. Park, A. P. Alivisatos, and P. L. McEuen., Appl.
Phys. Lett. 75, 301 共1999兲.
3
H. Park, J. Park, A. K. L. Lim, E. H. Anderson, A. P. Alivisatos, and P. L.
McEuen, Nature 共London兲 407, 57 共2000兲.
4
M. F. Lambert, M. F. Goffman, J. P. Bourgoin, and P. Hesto,
Nanotechnology 14, 772 共2003兲.
5
D. R. Strachan, D. E. Smith, D. E. Johnston, T. H. Park, M. J. Therien, D.
A. Bonnell, and A. T. Johnson, Appl. Phys. Lett. 86, 043109 共2005兲.
6
A. A. Houck, J. Labaizewicz, E. K. Chan, J. A. Folk, and I. L. Chuang,
Nano Lett. 5, 1685 共2005兲.
7
J. R. Black, Proc. IEEE 57, 1587 共1969兲.
8
D. Young and A. Christou, IEEE Trans. Reliab. 43, 186 共1994兲.
9
M. G. Burzo, P. L. Komarov, and P. E. Raad, IEEE 2002 Inter Society
Conference on Thermal Phenomena 共2000兲, pp. 142–149.
10
F. P. Incorpera and D. P. DeWitt, Introduction to Heat Transfer, 4th ed.
共Wiley, New York, 2002兲.
11
L. Lu, W. Yi, and D. L. Zhang, Rev. Sci. Instrum. 72, 2996 共2001兲.
1
Downloaded 21 Mar 2010 to 132.66.144.233. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp